Side, a long‑standing services provider formerly known as PTW, has entered into a six‑month memorandum of understanding (MOU) with Modl.ai, a Danish artificial‑intelligence start‑up that has been active in the games sector since its 2018 launch and secured $1.7 million in seed funding two years later. The partnership is designed to combine Side’s deep QA expertise with Modl.ai’s emerging AI capabilities to create a hybrid, AI‑enabled quality‑assurance platform for video‑game developers. Martin McBride, Side’s chief commercial officer, describes the collaboration as “bringing together the best of Side and the best of Modl.ai in a unique solution.” He adds that the industry has been searching for a "holy‑grail" method to augment QA services with AI, and that both companies believe they have finally found a workable answer. Earlier AI‑driven QA tools, including Modl.ai’s previous offering, required developers to ship full builds—or even source code—into a testing environment that relied on plugins, SDKs, and tight integration.
This approach created significant friction, as developers had to expose internal assets and maintain complex pipelines, a problem that founder Christoffer Holmgard admits “became difficult over time.” The new system eliminates that need for deep integration. Instead, developers deliver the same product they would hand to a human tester or eventual consumer, install it on a device that mirrors the target platform, and let the AI observe the game’s visual output, simulate inputs, and compile a comprehensive data set of every interaction.
The AI workflow begins with what Modl.ai calls “building game knowledge.” Developers upload assets, design documents, screenshots, and other relevant material, then converse with the AI in natural language to teach it the fundamentals of the title. Holmgard explains that this dialogue can include images and textual descriptions, allowing the system to form a contextual understanding of game mechanics, narrative beats, and visual cues. Once the AI has a baseline comprehension, the development team defines the testing objectives.
These can range from conventional test cases—such as verifying UI elements or level loading—to more exploratory scenarios that are specifically crafted to leverage the AI’s pattern‑recognition strengths. Holmgard notes that human expertise is still essential at this stage to shape prompts and guide the AI toward the right behaviours.
During a “run” (Modl.ai’s term for a testing session), the AI plays the game, captures screen recordings, logs input sequences, and records system telemetry. After the run, it produces bug reports that include textual descriptions, short video snippets as evidence, and full log files. These reports are then handed to a human QA specialist for review.
The specialist decides whether each finding represents a genuine defect, a design feature, or a false positive, and provides feedback that the AI uses to refine its future assessments. Distinguishing bugs from intended features remains a nuanced challenge.
Holmgard points out that the AI’s judgments improve over time through a combination of prompt engineering, pre‑built game knowledge, and continuous feedback from human testers. As the system receives corrective input—"this is a bug," "that is a feature," or "this is acceptable for this reason"—its internal models become better attuned to the particular design language of the game.
Side envisions a broader impact on the QA workforce. By automating repetitive, low‑value tasks such as smoke tests and regression checks, the technology frees human testers to focus on higher‑order analysis, strategic thinking, and creative exploration. McBride explains that this shift enables QA staff to evolve into "AI engineers," opening new career pathways, increasing compensation, and potentially repatriating jobs to regions where they have previously disappeared due to cost pressures.
He emphasizes that AI is not a replacement but a tool that moves people up the value chain, allowing them to stay longer in the industry and perform more fulfilling work. Benedikte Mikkelsen, Modl.ai’s chief product officer, echoes this sentiment, noting that many current tests are tedious and repetitive. By offloading those to AI, human testers can concentrate on assessing gameplay feel, balance, and creative edge—activities that truly add value.
McBride also stresses that AI will never fully replace human judgment because testing price points in certain low‑cost markets remain too low to justify a complete automation. Instead, the partnership aims to augment human capability, reduce error rates, and streamline workflows. He describes AI as a "sliding scale" where the proportion of automation can be adjusted, but a human will always remain in the loop—this is the core ethos of Side’s approach.
When asked about the types of games the solution currently supports, COO Oliver Deslandes admits that the technology is best suited for simpler titles—puzzle games, card games, and other experiences with repeatable mechanics and modest graphical intensity. Fast‑paced, 3D open‑world or fighting games present a greater challenge and are slated for later development phases.
Deslandes stresses that early adopters should be willing to pilot the system, understand its limitations, and collaborate on iterative improvements. In summary, the Side‑Modl.ai MOU represents a strategic effort to blend human expertise with machine intelligence in game QA. By removing the need for deep code integration, leveraging natural‑language interaction, and maintaining a human‑in‑the‑loop review process, the partnership aims to deliver faster, more accurate bug detection while elevating the role of QA professionals into higher‑skill, higher‑pay positions. The initiative is still in its pilot stage, focusing on less complex games, but both companies view it as a stepping stone toward broader applicability across the gaming industry.